PCA.test: PCA-based Gene-Gene Interaction analysis.
In GeneGeneInteR: Tools for Testing Gene-Gene Interaction at the Gene Level

Description Usage Arguments Details Value References See Also Examples

PCA.test performs a Gene-Gene Interaction (GGI) analysis by testing the interaction between the principal components of the two genes. With method="Std" PCA is standardized using standard deviation for each variable. With method="GenFreq", dataset is standardized using the standard deviation under Hardy-Weinberg equilibrium, as proposed in the snpStats Bioconductor package.

1	PCA.test(Y, G1, G2, threshold = 0.8,method="GenFreq")

`Y`	numeric or factor vector with exactly two different values. `Y` is the response variable and should be of length equal to the number of rows of `G1` and `G2` arguments (number of individuals).
`G1`	SnpMatrix object. Must have a number of rows equal to the length of `Y`.
`G2`	SnpMatrix object. Must have a number of rows equal to the length of `Y`.
`threshold`	(optional) numeric comprised in [0, 1] interval. `threshold` is the minimum percentage of inertia kept for each PCA.
`method`	(optional) character string for PCA method. Must be one of the following: "GenFreq", "Std" (See details).

In a first step, PCA.test performs a Principal Components Analysis on both G1 and G2 genes that are interpreted as matrices of allele counts. With the method="Std", the dataset is standardized using variables standard deviation, while dataset is standardized using standard deviation under Hardy-Weinberg equilibrium for method="GenFreq". Principal components are then retrieved to describe each dataset with user-defined inertia percentage (parameter threshold) and used in a logistic regression model. The consists is testing the significance of the interaction terms using a Likelihood Ratio Test (see Li et al. (2009)).

A list with class "htest" containing the following components:

`statistic`	The value of the statistic: the deviance of the anova test.
`p.value`	The p-value for the test.
`estimate`	A vector of the residual deviances.
`parameter`	The degrees of freedom of the chi-squared distribution of the test statistic.
`null.value`	the value of the deviance under the null.
`alternative`	a character string describing the alternative.
`method`	a character string indicating the method used.
`data.name`	a character string giving the names of the data.

J. Li et al. (2009) Identification of gene-gene interaction using principal components. BMC Proceedings, 3 (Suppl. 7): S78

GGI

1
2
3

data(gene.pair)
PCA.test(Y=gene.pair$Y, G1=gene.pair$G1,G2=gene.pair$G2,threshold=0.7,method="Std")
PCA.test(Y=gene.pair$Y, G1=gene.pair$G1,G2=gene.pair$G2,threshold=0.7,method="GenFreq")